Default variable details

Some of the debops.rabbitmq_server default variables have more extensive configuration than simple strings or lists, here you can find documentation and examples for them.

rabbitmq_server__config
RabbitMQ configuration options
rabbitmq_server__plugins
rabbitmq_server__accounts
rabbitmq_server__user_limits
rabbitmq_server__vhosts
rabbitmq_server__vhost_limits
rabbitmq_server__exchanges
rabbitmq_server__queues
rabbitmq_server__bindings
rabbitmq_server__feature_flags
rabbitmq_server__global_parameters
rabbitmq_server__parameters
rabbitmq_server__policies

rabbitmq_server__config 

The rabbitmq_server__*_config variables describe contents of the /etc/rabbitmq/rabbitmq.config configuration file. Each entry in the rabbitmq_server__*_config variables is a YAML dictionary with specific parameters:

name

Required. The name of an Erlang application to configure. Each application can contain a set of configuration options. Configuration options from multiple applications with the same name parameter are merged together.

state

Optional. If not specified or present, a given application configuration will be included in the finished configuration file.

If absent, a given application configuration will be removed from the configuration file.

If ignore, a given application entry is not evaluated by the configuration template. This can be used to conditionally enable or disable configuration sections.

comment

Optional. A string or YAML text block which will be added as a comment to the configuration section.

weight

Optional. A positive or negative number which will be used to affect the position of a given Erlang application within the configuration file. The higher the number, the more a given application section "weighs", and therefore it will be placed lower in the finished configuration file. If not specified, 0 is used by default.

options

A YAML list of configuration options for a given Erlang application. See RabbitMQ configuration options for more details.

Examples

---

# Create a basic set of Erlang applications used by RabbitMQ, based on the
# example configuration file:

rabbitmq_server__config:

  - name: 'rabbit'
    weight: 1

  - name: 'rabbitmq_management'
    comment: |
      RabbitMQ Management Plugin

      See https://www.rabbitmq.com/management.html for details
    options: []
    weight: 2

  - name: 'rabbitmq_management_agent'
    weight: 3

RabbitMQ configuration options 

RabbitMQ is written in the Erlang programming language, which is also used for its configuration. YAML, used by Ansible, does not provide enough data types to directly map them to the Erlang data types used in the RabbitMQ configuration file, therefore the configuration used by debops.rabbitmq_server focuses on description of the desired data types and conditional activation of the configuration sections. This means that simple values like strings, numbers, lists are mapped directly, however more complex configuration needs to be written in Erlang using YAML text blocks. The role tries to detect the value type automatically, but in some cases you might need to use the extended YAML dictionary syntax described below.

The role does not provide original configuration variables due to the issues with template generation (commented out options are not supported). You can find a reference RabbitMQ configuration file after the service installation, in the /usr/share/doc/rabbitmq-server/rabbitmq.config.example.gz file. An example rabbitmq.config file is also available online.

RabbitMQ configuration options are included in the options parameter of an Erlang application section (see rabbitmq_server__config for more details). The options parameter is a YAML list, each entry is a YAML dictionary. The dictionary keys are used as option names, and dictionary values are used as option values. You can specify simple options this way:

---

# Example of a set of simple RabbitMQ options
rabbitmq_server__config:

  - name: 'rabbit'
    options:

      # String
      - example_option: 'value'

      # Simple list
      - tcp_listeners: [ 5672 ]

      # Boolean value
      - reverse_dns_lookups: True

      # Numbers
      - vm_memory_high_watermark: 0.4

      # Raw Erlang code (note absence of }, at the end)
      - tcp_listeners: |
          [{"127.0.0.1", 5672},
           {"::1",       5672}]

If a given dictionary contains a name parameter, the configuration template will switch to a more verbose option interpretation, using known parameters:

name

The name of a given configuration option. Multiple entries with the same name are merged together, with the latter ones takim precedence over the former.

value

Required. A value to set for a given option. The value can be an YAML string, a list, number, boolean.

YAML text block is used to indicate a raw Erlang code which should be used as a value. The raw Erlang code should not end with any flow control Erlang characters (} or },), they will be added automatically by the role.

type

Optional. Specify the type of a given value to use. If the type parameter is not specified, the template will try to select one based on the YAML value type. Supported value types:

string: a quoted string, selected automatically if a YAML string is used as the value;
list: a list of values, selected automatically if a YAML list is used as the value;
number: an unquoted number, selected automatically if a YAML number or float is used as the value;
boolean: a boolean true/false value, selected automatically if a YAML boolean is used as the value;
bit-string: a bit string value with special quotation marks. Only YAML strings are supported at this time;
bit-list: a list of bit-strings with special quotation marks. Only YAML strings are supported at this time. if the value type is set as bit-string and a YAML list is set, the role should change to a bit-list type automatically;
raw: a raw Erlang expression, inserted in the finished configuration file as-is. The Erlang code should not end with Erlang flow control characters } or },, they will be added automatically by the role. if the value is specified using a YAML text block, the raw type should be selected automatically, based on the number of lines used in the value;

option

Optional. If specified, the configuration option will use this value for the name instead of name.

state

Optional. If not specified or present, a given option be included in the finished configuration file.

If absent, a given option will be removed from the configuration file.

If ignore, a given option entry is not evaluated by the configuration template. This can be used to conditionally enable or disable configuration options.

comment

Optional. A string or YAML text block which will be added as a comment to the configuration option.

weight

Optional. A positive or negative number which will be used to affect the position of a given option within the configuration file. The higher the number, the more a given option "weighs", and therefore it will be placed lower in the finished configuration file. If not specified, 0 is used by default.

Examples

---

# Example of a set of verbose RabbitMQ options
rabbitmq_server__config:

  - name: 'rabbit'
    options:

      # String
      - name: 'example_option'
        value: 'value'
        type: 'string'

      # Simple list
      - name: 'tcp_listeners'
        value: [ 5672 ]
        type: 'list'

      # Boolean value
      - name: 'reverse_dns_lookups'
        value: True

      # Numbers
      - name: 'vm_memroy_high_watermark'
        value: 0.4

      # Bit-string (result: '<<"bit-string">>')
      - name: 'bit_option'
        value: 'bit-value'
        type: 'bit-string'

      # Bit-list (result: '[<<".*">>, <<".*">>, <<".*">>]')
      - name: 'default_permissions'
        value: [ '.*', '.*', '.*' ]
        type: 'bit-list'

      # Raw Erlang code (note absence of }, at the end)
      - name: 'tcp_listeners'
        value: |
          [{"127.0.0.1", 5672},
           {"::1",       5672}]
        type: 'raw'

rabbitmq_server__plugins 

The rabbitmq_server__*_plugins lists can be used to enable or disable RabbitMQ plugins conditionally. You can find the available plugins on a given host by running the command:

rabbitmq-plugins list

Each list entry is either a RabbitMQ plugin name, or a YAML dictionary with specific parameters:

name: The name of a RabbitMQ plugin to manage.
state: Optional. If not defined or present, the plugin will be enabled. If absent, the plugin will be disabled.
prefix: Optional. Custom install prefix to a Rabbit.

Examples

Enable the RabbitMQ Management Console agent:

rabbitmq_server__plugins:

  - 'rabbitmq_management_agent'

rabbitmq_server__accounts 

The rabbitmq_server__*_accounts list variables can be used to manage RabbitMQ user accounts. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the rabbitmq_user Ansible module. Some more common parameters:

user or name: The name of a given user account.
state: Optional. If not specified or present, the user account will be created. If absent, the user account will be removed.
password: Optional. Plaintext password of a given user account. If not specified, the role will generate a random password and store it in the secret/rabbitmq_server/accounts/ directory on the Ansible Controller. See debops.secret Ansible role for more details.
tags: Optional. A string or a YAML list of tags assigned to a given account. Possible choices: management, policymaker, monitoring, administrator.
vhost: Optional. Name of the virtual host to which a given set of permissions should apply. If not specified, / vhost is used by default.
configure_priv, read_priv, write_priv: Optional. A regular expression which defines what resources on a given virtual host the user can configure, read from or write to. By default the ^$ regexp is used which means no permissions are given to any resources on a virtual host.

Examples

Create an administrator account and a regular user account:

rabbitmq_server__accounts:

  - name: 'admin_account'
    vhost: '/'
    tags: [ 'administrator' ]
    configure_priv: '.*'
    read_priv: '.*'
    write_priv: '.*'

  - name: 'user_account'
    vhost: '/'
    read_priv: '.*'
    write_priv: '.*'

rabbitmq_server__user_limits 

The rabbitmq_server__*_user_limits list variables can be used to configure RabbitMQ per-user connection limits using Ansible. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_user_limits Ansible module.

The available parameters:

user: Required. Name of the RabbitMQ user to configure.
node: Optional. Limit the user limits to a specific RabbitMQ node.
max_connections: Optional. The maximum number of connections that can be open to a given RabbitMQ user.
max_channels: Optional. The maximum number of channels that can be open to a given RabbitMQ user.
state: Optional. If not specified or present, the user limit will be created. If absent, the user limit will be removed.

Examples

Define limits for a specific user account:

rabbitmq_server__user_limits:

  - user: 'admin_account'
    max_connections: 1000
    max_channels: 100

rabbitmq_server__vhosts 

The rabbitmq_server__*_vhosts list variables can be used to manage RabbitMQ virtual hosts. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the rabbitmq_vhost Ansible module. Some more common parameters:

name: The name of a given virtual host. If not specified, the whole list entry will be used as the name (see examples).
state: Optional. If not specified or present, the virtual host will be created. If absent, the virtual host will be removed.
tracing: Optional. Enable message tracing in a given virtual host.
default_queue_type: Optional. Sets the default queue type for the virtual host using rabbitmqctl update_vhost_metadata --default-queue-type. Supported values on RabbitMQ 3.13+: classic, quorum, stream. When unset, RabbitMQ uses its built-in default (classic unless overridden on the cluster). Useful on RabbitMQ 4.x where quorum queues can be made the default without resorting to policies.

Examples

Create a set of virtual hosts:

rabbitmq_server__vhosts:

  - 'vhost1'

  - 'vhost2'

  - name: 'vhost3'
    state: 'absent'

  - name: 'vhost4'
    default_queue_type: 'quorum'

rabbitmq_server__vhost_limits 

The rabbitmq_server__*_vhost_limits list variables can be used to configure RabbitMQ virtual host limits using Ansible. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_vhost_limits Ansible module. Available parameters:

vhost: Required. Name of the RabbitMQ virtual host to configure. The default virtual host is named /.
node: Optional. Limit the virtual host limits to a specific RabbitMQ node.
max_connections: Optional. The maximum number of connections that can be open to a given RabbitMQ virtual host.
max_queues: Optional. The maximum number of queues that can be created in a given RabbitMQ virtual host.
state: Optional. If not specified or present, the virtual host limit will be created. If absent, the virtual host limit will be removed.

Examples

Define virtual host limits for the default vhost:

rabbitmq_server__vhost_limits:

  - vhost: '/'
    max_connections: 1000
    max_queues: 100

Reset limists on the default vrrtual host:

rabbitmq_server__vhost_limits:

  - vhost: '/'
    state: 'absent'

rabbitmq_server__exchanges 

The rabbitmq_server__*_exchanges list variables can be used to manage RabbitMQ exchanges. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_exchange Ansible module.

List of supported parameters:

name: Required. The name of a given RabbitMQ exchange.
vhost: Optional. Specify the RabbitMQ virtual host to which a given exchange applies.
exchange_type: Optional. The type of a given RabbitMQ exchange. Supported choices: direct, fanout, topic, headers, x-consistent-hash, x-delayed-message, x-random, x-recent-history.
durable: Optional. If not specified or True, the exchange will be durable. If False, the exchange will be transient.
auto_delete: Optional. If not specified or False, the exchange will not be deleted when the last queue is unbound from it. If True, the exchange will be deleted when the last queue is unbound from it.
internal: Optional. If not specified or False, the exchange will be a regular exchange. If True, the exchange will be an internal exchange, only available for other exchanges.
arguments: Optional. A YAML dictionary with additional arguments to set for a given exchange.
state: Optional. If not specified or present, the exchange will be created. If absent, the exchange will be removed.

Examples

Create a set of RabbitMQ exchanges:

rabbitmq_server__exchanges:

  - name: 'exchange1'
    exchange_type: 'fanout'
    durable: False
    auto_delete: True

  - name: 'exchange2'
    exchange_type: 'topic'
    durable: True
    auto_delete: False

rabbitmq_server__queues 

The rabbitmq_server__*_queues list variables can be used to manage RabbitMQ queues. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_queue Ansible module.

List of supported parameters:

name: Required. The name of a given RabbitMQ queue.
vhost: Optional. Specify the RabbitMQ virtual host to which a given queue applies.
durable: Optional. If not specified or True, the queue will be durable. If False, the queue will be transient.
auto_delete: Optional. If not specified or False, the queue will not be deleted when the last consumer is removed. If True, the queue will be deleted when the last consumer is removed.
auto_expires: Optional. The time in milliseconds after which the queue will be deleted if it is not used.
dead_letter_exchange: Optional. The name of a dead-letter exchange to which messages will be republished if they are rejected or expire.
dead_letter_routing_key: Optional. The routing key to use when republishing messages to the dead-letter exchange.
max_length: Optional. The maximum number of messages that the queue will hold.
max_priority: Optional. The maximum priority of messages that the queue will hold.
message_ttl: Optional. The time in milliseconds after which a message will be removed from the queue if it is not consumed.
arguments: Optional. A YAML dictionary with additional arguments to set for a given queue.
state: Optional. If not specified or present, the queue will be created. If absent, the queue will be removed.

Examples

Create a set of RabbitMQ queues:

rabbitmq_server__queues:

  - name: 'queue1'
    durable: False
    auto_delete: True
    state: 'present'

  - name: 'queue2'
    durable: True
    auto_delete: False
    state: 'present'

rabbitmq_server__bindings 

The rabbitmq_server__*_bindings list variables can be used to manage RabbitMQ bindings. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_binding Ansible module.

List of parameters:

name: Required. The name of a given RabbitMQ exchange which will be the source of a given binding.
destination: Required. The name of a given RabbitMQ queue or another exchange which will be a destination of a given binding.
destination_type: Required. The type of a given destination. Supported choices: queue, exchange.
vhost: Optional. Specify the RabbitMQ virtual host to which a given binding applies.
routing_key: Optional. The routing key to use when binding a queue to an exchange.
arguments: Optional. A YAML dictionary with additional arguments to set for a given binding.
state: Optional. If not specified or present, the binding will be created. If absent, the binding will be removed.

Examples

Create a set of RabbitMQ bindings:

rabbitmq_server__bindings:

  - name: 'exchange1'
    destination: 'queue1'
    destination_type: 'queue'
    routing_key: 'example'
    state: 'present'

  - name: 'exchange2'
    destination: 'exchange1'
    destination_type: 'exchange'
    state: 'present'

rabbitmq_server__feature_flags 

The rabbitmq_server__*_feature_flags list variables can be used to manage RabbitMQ feature flags. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_feature_flag Ansible module.

Supported parameters:

name: Required. The name of a given RabbitMQ feature flag.
node: Optional. The name of a RabbitMQ node to which a given feature flag applies.

Examples

Enable the maintenance_mode_status feature flag on a specific Erlang node:

rabbitmq_server__feature_flags:

  - name: 'maintenance_mode_status'
    node: 'rabbit@node1'

rabbitmq_server__global_parameters 

The rabbitmq_server__*_global_parameters list variables can be used to manage RabbitMQ global parameters, defined for the entire cluster. Each list entry is a YAML dictionary with specific global parameters. The parameter names are the same as the community.rabbitmq.rabbitmq_global_parameter Ansible module.

Configuration entry parameters:

name: Required. The name of a given RabbitMQ parameter being set.
value: The value of a given parameter in a JSON format. The values are usually quoted using single quotes and contain double-quotes.
node: Optional. The name of a RabbitMQ node to which a given parameter applies.
state: Optional. If not specified or present, the parameter will be created. If absent, the parameter will be removed.

Examples

Set the value of the cluster_name global parameter:

rabbitmq_server__global_parameters:

  - name: 'cluster_name'
    value: '"my-cluster"'
    state: 'present'

rabbitmq_server__parameters 

The rabbitmq_server__*_parameters list variables can be used to manage RabbitMQ parameters. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the rabbitmq_parameter Ansible module. Some more common parameters:

component: Required. Name of the component of which the parameter is being set.
name: Required. The name of a given RabbitMQ parameter being set.
value: The value of a given parameter in a JSON format. The values are usually quoted using single quotes and contain double-quotes.
vhost: Optional. Specify the RabbitMQ virtual host to which a given parameter applies.
state: Optional. If not specified or present, the parameter will be created. If absent, the parameter will be removed.

Examples

Define a RabbitMQ parameter:

rabbitmq_server__parameters:

  - component: 'federation'
    name: 'local-username'
    value: '"guest"'

rabbitmq_server__policies 

The rabbitmq_server__*_policies list variables can be used to manage RabbitMQ policies. Each list entry is a YAML dictionary with specific parameters. The parameter names are the same as the rabbitmq_policy Ansible module. Some more common parameters:

name: Required. The name of a given RabbitMQ policy.
pattern: Required. A regexp pattern of RabbitMQ queue names to which a given policy applies.
tags: Required. An YAML dictionary with key/value parameters that describe the policy. Relevant documentation can be found in the RabbitMQ Management Console, Admin section, Policies.
vhost: Optional. Specify the RabbitMQ virtual host to which a given policy applies.
apply_to: Optional. The resource type to which a given policy applies to. Supported choices: all, exchanges, queues. If not specified, all is used by default.
state: Optional. If not specified or present, the policy will be created. If absent, the policy will be removed.
priority: Optional. The numerical priority of a given policy, used for sorting.

Examples

Create a set of RabbitMQ policies:

rabbitmq_server__policies:

  - name: 'HA'
    pattern: '.*'
    tags:
      'ha-mode': 'all'